Faults in a 110 MW Generator 1

[Addy71]

Currently overhauling a 110 MW generator at client’s site. Major observations as under:
1. Rotor has an interturn short – impedance is reduced by 17 % compared to the factory values. Surge waveform shows separation when tested from either end (+ve & -ve). Cleaning did not help.

2. Generator shaft is magnetized, as are the centering rings of the endbells.

3. Bearing insulation resistance was 2 megaohms.

4. Exciter end bearing shows electrical pit marks, but not in the load zone.

5. About 20 grub screws used to lock the rotor wedges were missing. Located 14 so far, some stuck in the stator ventilating ducts. Many rotor wedges were thus loose.

6. Boroscope inspection under the retaining rings did not reveal any abnormality.

7. One rotor coil has shifted axially leading to partial blockage of some of the ventilating holes.

This is a 7 year old gas turbine generator which is started every day morning, run at full load & shut down at night (first time I’ve seen this).

Answers/opinion sought:
1. How often do you run generators with rotor interturn shorts? I have seen two earlier, but with minor impedance reduction (surge test would show separation of the waveforms).

2. At what percentage of impedance reduction would say that a repair is a must?

3. Is there any way in which the shaft magnetization & interturn short are related?

4. I know IEEE specifies minimum bearing insulation as 5 megaohms. However, I can’t see the shaft voltage jumping 2 megaohms of insulation. Also, a shaft grounding brush is in place. So, how did the shaft get magnetized & the bearing get the pitting marks?

5. Client is not interested in repairing the fault or demagnetizing the shaft. What implication would a magnetized shaft have on reliability?

Any answers or feedback would be most appreciated.

Regards,

Aditya

 

[edison123]

Aditya

Cyclical loading the worst possible load on the generator due to frequent differential expansion between copper and insulation.

As for rotor inter-turn short, how was the generator vibration before overhaul ? Severe rotor turn shorts lead to excessive vibrations due to uneven magnetic pull. So more than the % impedance unbalance, it's the vibration that would determine a costly rotor rewind, I guess.

Aren't the retaining rings supposed to be non-magnetic ? Those grub screws on the rotor wedges working loose is a serious issue since they could damage a healthy stator winding. Who is the OEM of this generator anyway ? Sounds like poor workmanship and poor choice of materials.

I have seen bearing insulation as low as 1 megohm working fine, though a higher value is obviously preferred. If the isolation is a pad under the bearing pedestal, you could replace with the new one, which is not very costly.

How to demagnetize the shaft ? I too would like an answer to that.

Muthu

http://www.edison.co.in

 

[Addy71]

Muthu,

The generator vibration shows a rising trend, from 3.0 mm/sec, RMS at commissioning to 8.5 mm/sec, RMS now. However, the vibration levels have been steady over the last six months.

The retaining rings are non-magnetic, however the centering rings are not. The centering rings are used to position the retaining rings. In some designs, both come off together, in others they are separate. The generator is BHEL

We are changing the bearing insulation. Wonder what would happen if I insulate the second bearing also? Would it mess up the turbine bearings?

Shaft can be demagnetized. Haven't done it myself, but I believe you pass DC current & try to reverse the dipoles. Spoke with Aron at Hardwar, says its a normal procedure. Client is not willing though.

Regards,

Aditya

 

[edison123]

Ah, centering rings, my bad. But I wonder what prevents OEM to go non-magnetic there too.

If 8.5 mm/sec rms, loose grub screws, axial movement of the rotor winding and turn shorts don't convince the "client" to go for rewind, I guess, nothing else will.

Thx for the info on shaft demagnetization. Need to see one.

BTW, when you say surge test, you mean RSO (TDR) and not surge comparison test (Baker instrument), right ?

IMO, the probability of problem being under the retaining rings is high. Do you plan to remove the RR's during the overhaul ?

Muthu

http://www.edison.co.in

 

[Addy71]

Muthu,

Till date, I had never paid attention to whether centering rings are magnetic or not. Noticed it this time only when it grabbed a screwdriver! Cost saving by BHEL, maybe?

RSO & surge were both done & gave the same results. PTA has popularized RSO in India; however CPRI, ERDA & ABB have always preferred surge. I too prefer the surge test on the rotor as I think the low voltage of the RSO will not pick up resistive faults.

We spent two shifts inspecting under the RRs with a boroscope. Could not see anything. Knuckles are okay, no elongation. Insulation & blocks too are intact. Client in a big hurry to box it up, says we'll remove the rings next time to repair the fault.

Have you repaired similar faults? At what values/patterns did the client decide to go for repair?

Regards,

Aditya

Regards,

Aditya

 

[rmw]

Edison,

Unfortunately, cycling is the nature of the beast for GT units. That is why they are favored for peaking applications. I have seen units with higher start counts than actual operating hours. It is good for the repair folks; know any? Hee hee. This one actually sounds like it gets to run once it is started in the morning. I know of GT units that are started as often as 3 times per day.

Addy,

Is there any evidence of hot spots on the RR's? This rotor is not going to heal itself and this client is about to learn the pay me now or pay me later rule. Especially in light of the nature of the duty it is asked to do.

If you isolate the second bearing, it will force the shaft currents to go to ground through the turbine bearings if the grounding brushes fail to do their thing for a variety of reasons. The turbine people get real testy about that. If this is an aeroderivative turbine utilizing ball bearings, it could have a real negative impact. If it wasn't designed that way, don't change it.

rmw

 

[edison123]

Addy

How did you do the surge without removing the RR ? Did you use the shaft as the return path ?

About 2 years back, I got a TG rotor (only 16 MW) with an earth fault. It had RR with holes drilled on it (apparently for cooling - strange). We pulled the RR on exciter end and found a clean burn through hole under RR insulation with a few turns short in the upper layer of the coil under the burn through. Did a surge check under that condition and got a double wave. Cleaned up the insulation of the turns carefully and the short vanished. Provided new under band insulation and got the IR up to 700 Meg. Machine running still.

rmw - Pay me now or you can pay me later. Can live with that.

Muthu

http://www.edison.co.in

 

[Addy71]

Muthu,

We applied the surge from either end, viz. +ve & -ve and superimposed the resultant waveforms.

Interesting case, the 16 MW. So basically, the earth fault was gone the moment you took the ring off but the short persisted and showed up in the surge. Do you have a record of the rotor impedance before & after the insulation clean-up?

RMW,

There are more issues as discussed with EPete on the other forum (hats off to him!). If the interturn fault is causing shaft currents (not confident about that yet), bearing problems will re-occur.

They've reported accessory gearbox failures beacuse of shaft current. That is another mystery as the gearbox is beyond the turbine & the turbine bearings are not insulated. How come those bearings didn't fail? Seems the gear teeth would get electrical pitting marks, as detected from vibration. They had replaced the generator bearing insulation then.

For VFD motors, you HAVE to insulate both bearings or you have a problem. For generators & large motors with a single side bearing, again the bearing is insulated. How come turbine bearings don't have problems in those cases?

Regards,

Aditya

 

[edison123]

Aditya

Impedance at 33 deg C before/after - 8.7 / 9.9 ohms

Resistance at 33 deg C - 0.34 ohms

But rotor came to us for earth fault problem. We noted the turn shorts only after removing the RR. And Genny was 6.6 KV, 12.8 MW, not 16 MW.

Muthu

http://www.edison.co.in

 

[glendale]

The magnetisation of the shaft will be due to unbalanced flux in the rotor, typically caused by a double earth fault. The double earth fault shorts out part of one poles winding. The shorts may not be obvious at rotor standstill and so may be difficult to find and repair.

I de-magnetised a rotor (100MW two pole cylindrical) by wrapping 50 turns of welding cable around the length of the rotor, then passed 100A DC from a welding set through the turns (slowly decreasing the current). Then reverse the polarity and repeat reducing the current by 20% each time. It might sound a bit non-technical but I was in the middle of nowhere at the time! It worked anyway.

 

[Addy71]

Glendale,

Thanks for the input, this is great. Did you wrap the cable around the shaft or rotor bore or both? How long did you pass the current before reversing the polarity? Did you use a gauss meter to assess the magnetism level?

Regards,

Aditya

 

[glendale]

I wrapped the turns along the length of the rotor including the shaft. The current was held on for just a few seconds. As soon as I reached the maximum current I then started to reduce it, the cycle lasted about 10 seconds.
I didn't have a gauss meter on site and measured the effectivness of the de-gaussing by hanging a series of heavy steel washers off the shaft, at each change of polarity less washers could be supported. At the end of the de-gaussing the shaft was completely de-magnetised. I know this sounds rather un-seientific, but thats how I had to do it with the equipment available!!

 

[Addy71]

Glendale,

Many thanks again for the input. Last couple of questions - how long did the whole process take & what was the voltage applied - one volt, some volts or higher?

Will not bother you further after this one.

Regards,

Aditya

 

[edison123]

Thanks Glendale for a valuable LPS tip.

I'm also interested to know how you went about it.

Muthu

http://www.edison.co.in

 

[rmw]

Glendale,

Could you write it up in sufficient detail and them make it a FAQ for this forum?

rmw